In recent years colloidal metal particles have been extensively used in biological, biochemical and immunological investigations. Especially in the field of immunological assays, the application of colloidal metal particles for labeling of any type of biomolecule has been very successful.
For the production of colloidal metal particles, in particular colloidal gold particles, many different methods have been described. Colloidal gold particles are usually obtained by the reduction of solutions of tetrachloroauric acid, and relevant biomolecules are bound to these particles by adsorption in one or several further steps. The size of the colloidal metal particles can be controlled specifically by means of the initial concentration of the different reaction partners (G. Frens, Nature Physical Science 241:20-22, 1973). Variation of particle size can also be optimized by specific selection of the production procedure as described in EP B 426,300.
The binding of biomolecules such as, for example, antibodies, lectins or even nucleic acids, is mainly adsorptive, with the respective loading conditions strongly depending on the physico-chemical properties, e.g., isoelectric point, of the biomolecule. Principal reflections on the coating of colloidal metal particles can be found in the relevant technical literature, e.g., deMey, “The Preparation and Use of Gold Probes,” in Immunocytochemistry, J. M. Polak and S. V. Noorden, pp. 115-145, Wright, Bristol, 1986 and G. T. Hermanson, “Preparation of Colloidal-Gold-Labeled Proteins,” in Bioconjugate Techniques, pp. 593-604, Academic Press, 1996.
It has been proven to be necessary and useful to stabilize colloidal metal particles coated with biomolecules by adding further appropriate substances. The addition of such stabilizing and blocking substances produces the effect that probably non-saturated, “sticky” sites on the particles are blocked. In addition, by these substances the aggregation of the particles is minimized, and the re-diffusion of the relevant biomolecule is reduced.
State of the art stabilizers are, among others, inert proteins such as, for example, bovine serum albumin or water-soluble technical polymers such as polyethylene glycol (molecular weight of 20,000 Da), polyvinyl pyrrolidone, polyvinyl alcohol, polyvinyl sulfate, dextran as well as gelatin, cf, e.g., De Mey, supra; Beesly, J. E., “Colloidal Gold: A New Perspective for Cytochemical Marking,” Microscopy Handbooks 17, Oxford University Press, 1989, in particular pp. 1-14; Behnke, Eur. J Cell. Biol. 41: 326-338, 1986; DE 2420531 C3; and Meisel et al., J. Phys. Chem. 85:179-187, 1981.
The direct coating of biomolecules on the surface of colloidal metal particles has, however, also some important disadvantages. Such a disadvantage is, for example, that successful methods cannot be transferred easily from one biomolecule to a second, and that the stability of such particles has proven to be problematic (cf. de Mey, supra, p. 116 and pp. 122-123).
To avoid such problems of the older state of the art, some new methods have been developed in recent years. In particular the so-called core-shell particles are to be mentioned here. Core-shell particles consist of a (usually) metal core and of a shell surrounding the particle completely. Important examples of such shell substances are, for example, bovine serum albumin (BSA), described in EP 258,963 and EP 428,412, denatured proteins such as, for example, gelatin, water-soluble polymers such as, e.g., aminodextran (U.S. Pat. No. 5,248,772) and shells made of low-molecular thiol compounds or thiolized proteins (EP 489 465).
To stabilize the shell, the shell substances are usually cross-linked, e.g., U.S. Pat. No. 5,248,772. Such shells contain functional groups to which the desired biomolecules can bind in a covalent way as it is generally known.
The covalent coupling of the biomolecules to the core-shell particles has some advantages over the direct adsorptive binding to the colloidal surface. By covalent coupling, bleeding of the biomolecules can be avoided. Furthermore, unspecific interactions are stopped which are provoked by the remaining free surface areas inaccessible for adsorption during the adsorptive loading procedure. A post-treatment of the surface by adding stabilizers is not necessary with core-shell particles. By the use of a well-soluble hydrophilic shell, the stability of the resulting core-shell-protein conjugates can be remarkably improved. Instability caused by the aggregation of the conjugates, as it is repeatedly observed, with adsorptively loaded colloidal-protein conjugates can thus be avoided. Furthermore, covalent coupling of the proteins to the core-shell particles is to a large extent independent of the physico-chemical properties of the proteins, so that even proteins can be used which are not suitable for an adsorptive binding.
The shell substances used in the state of the art adsorb to the metal surface by complex and hitherto not completely understood mechanisms involving electrostatic as well as hydrophobic interactions. The type of shell substance determines which bond type prevails when the shell substances bind to the colloidal metal particles.
Interactions of metals, in particular gold, with compounds containing sulfur lead to a reinforced adsorptive binding. EP 369,515 describes the coating of colloidal gold particles with low-molecular compounds containing thiol. By additional functional groups, the relevant biomolecules can be bound to this thiol shell.
It is also known from the state of the art that chemically cross-linked aminodextrans can be used as the shell substance (U.S. Pat. No. 5,248,772). Cross-linking of aminodextran is required to give sufficient stability to the particles produced in such a way. A further disadvantage of the method described in U.S. Pat. No. 5,248,772 is that SH groups suitable for coupling must be inserted into the shell of cross-linked aminodextran afterwards. Such a procedure requires several additional working steps. In addition, the insertion rate of the SH groups can only be controlled less exactly, and the analysis of the insertion efficiency is almost not possible.